Whatever social media platform you look at, it's increasingly seeming practically impossible to avoid AI-generated content entirely. In recent months, we've seen many platforms roll out tools to help users weed out the slop from the authentic content. At least in LinkedIn's case, these tools are already proving extremely popular with users.
Over a million people have now clicked on LinkedIn's 'seems like AI slop' button, which rolled out earlier this August, according to Hari Srinivasan, the Chief Product Officer at LinkedIn.
The tool allows people to report content they suspect to be AI-generated by clicking the option in the three-dot menu on the top right of any post or comment. If a post receives enough community feedback, creators get a message in their Post Analytics to let them know that their content is perceived to be AI.
Srinivasan estimates that LinkedIn members are overall now experiencing 40% fewer views of what it classifies as AI slop than just a few weeks ago, though he caveats that everyone has a different feed, so that exact ratio will look different for everyone.
LinkedIn was at one point considered to be one of the platforms worst hit by AI slop. In July, AI detection firm Pangram found that over 40% of longform posts were flagged as fully AI-generated.
Detecting AI slop is never a sure thing, as most AI detectors can be unreliable, but it could be getting easier soon. Some of the most popular AI tools, such as Anthropic's Claude, are incorporating 'watermarking' technology, which will leave outwardly undetectable signs in any text it generates that can later indicate if it's AI-generated.
Plenty of other social media platforms have been rolling out slop detection tools. At the start of this August, Snapchat rolled out an integration with Pangram to mark suspected AI-generated text, while also removing AI-generated videos from its Spotlight tab. Spotify also rolled out new AI Persona labels to help users spot AI-generated music.